Interactive Functional Programming

We propose a new kind of execution environment where applications can be debugged and re-programmed while they are being used. We call our overall concept interactive programming. We develop some of the key components of interactive programming in the setting of a pure, call-by-value functional language. We illustrate our ideas via a proof-of-concept implementation called lambdaCalc, but leave several important components of the overall vision, including efficient incremental update and scaling to large programs, for future work. Our specific achievements are as follows. First, we show how to reify the execution of a program into a live document which can be interactively decomposed into both sequential steps and parallel slices. We give a novel characterisation of forward and backward dynamic slicing and show that for a fixed computation the two problems describe a Galois connection. Second, we introduce a novel execution indexing scheme which derives execution differences from program differences. Our scheme supports the wholesale reorganisation of a computation via operations such as moves and splices. The programmer is able to see the consequences of edits on the intensional structure of the execution. Where possible, node identity is preserved, allowing an edit to be made whilst an execution is being explored and the changes to be reflected in the user's current view of the execution.